By now it’s conventional wisdom: screen for cancers early and often in order to catch potentially deadly cases before they spread.
But for the last four years a government-elected panel of doctors has been considering the question of if and when men should be screened for prostate cancer in the United States. It is currently the second most commonly diagnosed form of cancer—and many in the medical world believe that that rate of diagnosis is unnecessarily and dangerously inflated.
The culprit under investigation by the government panel was a fairly routine blood test that screens for elevated levels of the protein PSA, which sometimes indicates prostate cancer but often does not. Developed in the late 1980s, the test became integrated into standard medical exams in the 1990s, a change that many doctors credit with significantly reducing the number of prostate-cancer-related deaths in the United States; it was frequently championed by doctors, advocates, and various celebrities.
The panel had previously advised doctors to perform the screening only in limited cases. On Monday, in an article published online in the Annals of Internal Medicine, the members of the panel announced their decision to carry their previous disclaimers even farther. They stamped a D grade on PSA screening, explicitly discouraging its use.
The panel’s recommendation was based on findings suggesting that widespread PSA screening too easily leads to aggressive and unnecessary interventions, both by turning up false-positive results, and by alerting patients to tumors that are not, in fact, life threatening. The treatments for prostate cancer, including surgery and radiation, can result in serious, life-changing side effects, including impotence and incontinence, as well as more critical complications, including death.
The panel’s recommendation is not directed at the worth of the PSA test itself, but rather of its use without specific reason for concern. “A screening test is something you do on someone who already feels fine,” said Dr. Virginia Moyer, the chair of the U.S. Preventive Services Task Force. “You can only make them worse.”
“Indeed, it is still conceivable that prostate-cancer screening kills more people than it saves,” says Otis Brawley, the chief medical officer of the American Cancer Society, who wrote an editorial in the Annals of Internal Medicine that accompanied the government panel’s report. (Brawley himself was not a member of the panel.) PSA screening increases the risk of depression and suicide, he says, and even though he believes there is still no flawless clinical trial on the benefits or downsides of the PSA screen, Brawley also thinks that “the known harms are greater than the benefits.”
According to the panel, which looked at two large field trials that each followed several thousand men over more than 10 years, a significant number of PSA-detected prostate tumors either will not continue to grow or will grow so slowly that patients will likely die of other causes before they die from their untreated prostate cancers. “There is convincing evidence that the number of men who avoid dying of prostate cancer because of screening after 10 to 14 years is, at best, very small,” the members of the panel wrote in their decision.
The panel concluded that PSA screening leads to an over-diagnosis rate of between 17 and 50 percent, though a more exact number, they wrote, was difficult to pin down.
The panel’s conclusion exposes a major fault line in American medicine: the tension between the benefits of thorough screening and the dangers of over-detection. The new recommendations against PSA screening were greeted with indignant incredulity by many in the medical community. “It is stunning that you recommend against screening for any patients,” said Dr. Peter Schlegel, chair of the department of urology at Weill Cornell Medical school in New York.
Schlegel pointed to problems in both of the two largest studies taken into consideration by the panel, one study conducted in America, the other in Europe. The American trial turned up no evidence that PSA screening helped to prevent death from prostate cancer, but that, Schlegel says, is the result of the trial’s design. “The basic problem is that they compared PSA screening to usual care,” Schlegel said. “In the American study, you’re comparing PSA screening to screening, because screening is involved in the usual care ... I don’t know how the panel could have overlooked this,” he said.
In the European trials, which did reflect a slight but significant drop in mortality rates, there still remained the problem of the timeline, Schlegel said: the European study followed their subjects for a median time of nine years. But prostate cancer can take 10 to 15 years to spread to a lethal stage.
“It is … conceivable that prostate cancer screening kills more people than it saves,” says Otis Brawley, the chief medical officer of the American Cancer Society.
Moyer, chair of the panel, said these critiques were misleadingly selective. The European trial actually consisted of five individual subsets, done in five countries. Two of the five, said Moyer, turned up some indication of a benefit from PSA screening, but taken as a whole, they did not. “If you cherry pick and look only at the most optimistic studies, you do see benefit—and in fact, we took that into account in our recommendations ... If you look at all of the studies together, what you see is the best estimate is no benefit.”
Schlegel and others also objected to the absence of any urologists or cancer specialists on the panel.
Moyer, addressing this concern, said that, urologists and oncologists had been deeply involved in consultation, but that the panel’s guidelines were designed for preventative care, not for those doctors actually involved in the treatment of a disease.
Screening for PSA levels is not the same thing as testing for prostate cancer. PSA levels increase with age and are tied to enlargement of the prostate, but not necessarily indicative of prostate cancer. Once a PSA test comes back with an abnormal result, a biopsy is required to make a diagnosis. Even when cancer is detected, it is unclear whether treatment for the slow-moving form of the disease is always warranted.
As a result, the problem may not be so much with the test, but the ways in which doctors have used those results in treating patients, says David Samadi, vice chair of the department of urology at Mt. Sinai Medical Center in New York. Samadi called the panel’s ruling “aggressive” and “irresponsible.” He said doctors need to know how to understand the meaning of an individual patient’s PSA level—based on age, ethnicity, family background, and other risk factors. “You need to have that [PSA] number, but then what you do with that number is really in the hands of that doctor,” Samadi said.
Dr. Paul Knoepfler is both a cancer biologist at UC Davis and a survivor of prostate cancer. He was diagnosed two and a half years ago, at the unusually young age of 42. At the time his PSA screening revealed only a slightly higher level of the antigen than the cut-off for his age group. But he went forward with a painful biopsy, and then opted for surgery when the biopsy showed cancer.
He says he can see both sides of the argument over routine PSA screening. From the perspective of having been a patient, he knows that “once you get that abnormal PSA response, most men just feel like they have to get treatment,” he says. “How many men can realistically sit back and not do anything? It’s like a switch. Once it's thrown, they have to go and do something.”
Refining the way in which the screening is administered would be preferable to discarding it altogether, Knoepfler believes. “PSA is best used if it is evaluated relative to a man’s age,” he says.